Remote Healthcare
Working Memory Test
By Alex Pech and Nathan Sherburn
Supervised by John Heasman
Overview
Background
Working memory has been found to play a significant role in hearing
performance. This finding is changing our current theoretical models
of listening performance in both hearing aid and cochlear implant
recipients. It is hypothesised that, while surgery and the implant
technology play an extremely important role in a cochlear
implantee’s hearing performance, a more holistic view (one that
includes the recipients ability to process information) may provide a
deeper understanding and uncover new ways to improve implant
performance.
Design
Our Research
This project is centred around the hypothesis that as the signal to
noise ratio (SNR) increases, people are forced to spend more time
processing the information using their working memory. This, in turn,
leads to a decrease in the amount of information actually stored.
Technology
Remote Healthcare
Platform
The working memory tests used in this application are each based
on the digit span test. The digit span test requires participants to
listen or watch as a string of digits is presented to them before
repeating the digits back in reverse order.
A number of different application platforms were considered for
deployment including Android, iOS, Windows and OSX. Deploying
the test as a web application however provided the advantage of
being OS independent and therefore increasing the number of
people who would be able to use the software.
75%
Given the simple nature of these digit span-based tests, it is possible
to allow participants to self-administer the examination. This makes
the testing process far easier for participants as they do not need to
travel to and from the testing clinic. It also makes it possible to
collect larger amounts of data.
50%
Digit Span Test
25%
The digit span test was chosen for a number of reasons including:
100%
0%
quiet
Software
A great deal of the software used in this project came from open
source initiatives such as jQuery, mongoDB, node js and express.
These provided a robust scaffold for the application and reduced
production time significantly.
• the ease of implementation from a technical stand point
• the minimal time required to explain the tests to subjects
Processing
Storage
noise
In order to test this hypothesis we created a self-administered,
working memory test application. The application requires users to
complete visual, listening and listening with noise working
memory tests.
It is our hope that this application will provide a platform to help
researchers answer questions such as “Does the amount of
processing required decrease as implantees become accustomed
to their implants?” and “Why do some people produce good listening
test scores but still report difficulties hearing in the real world?”.
• the wider range of cognitive abilities who are able to partake
Psignifit
• the minimal dependency on language skills
In order to analyse the
information provided by the
user, the open source,
psignifit toolbox, was used.
This toolbox allowed us to
create a model of the user’s
hearing performance in
various signal to noise ratio
(SNR) conditions. The
model’s parameters were
then analysed to determine
the goodness of fit.
This testing platform has the potential to obtain powerful
performance metrics which may reshape the way cochlear implant
technology works. It will allow researchers and implantees to track
their performance over time and learn more about the role working
plays in hearing.
The Design
The test is broken into two stages. The first stage consists of a digit
discrimination hurdle and a 100% intelligibility in noise
measurement. The digit discrimination ensures the participants are
able to hear all the numbers with no background noise while 100%
intelligibility in noise measurement determines the maximum level of
background noise in which the participant is still able to hear every
number.
Digit Span
(Visual WM)
Digit Span
(Listening WM)
Digit Span
(Listening w/
Noise WM)
The second stage consists of three working memory tests. The first
test is a visual working memory test which provides us with a
baseline working memory capacity. The listening and listening in
noise working memory tests will then stress the participants working
memory in the auditory domain.
Results
References
Industry Value
Student Learnings
Future Possibilities
The finished prototype is able to capture
data from any internet connected device
capable of running a modern browser. This
data can then be stored in a central
database where audiologists, engineers
and clinicians are able to access and
analyse it.
Both students learnt a great deal about
web application development, interaction
design, project management, cochlear
implant technology and audiology.
A successful trial of this project may lead to
new results and answer some of the most
difficult questions in audiology.
Plans have been made to roll out a trial run
in April 2015.
These learnings, while interesting in and of
themselves, will, no doubt, come in very
useful for future projects undertaken by the
students.
These results would, in turn, improve both
pre-operation outcome predictions and
post-operation rehabilitation for cochlear
implant recipients.
Woods DL, Kishiyama MM, Yund EW, et al. Improving digit span assessment of short-term verbal
memory. Journal of clinical and experimental neuropsychology. 2011;33(1):101-111.
doi:10.1080/13803395.2010.493149.
Burkholder R, Pisoni D. Digit span recall error analysis in pediatric cochlear implant users. International
congress series / Excerpta Medica. 2004;1273:312-315.
Smits C, Theo Goverts S, Festen JM (2013) The digits-in-noise test: assessing auditory speech
recognition abilities in noise. J Acoust Soc Am 133: 1693–1706
Geers AE, Pisoni DB, Brenner C. Complex Working Memory Span in Cochlear Implanted and Normal
Hearing Teenagers. Otology & neurotology : official publication of the American Otological Society,
American Neurotology Society [and] European Academy of Otology and Neurotology. 2013;34(3):396401. doi:10.1097/MAO.0b013e318277a0cb.